# SEO Francisco Claude Skill Pack

Version: 2026-07-03

These are reusable skill specs you can adapt into Claude projects, Claude Code skills, or your own agent instructions. Each skill is intentionally narrow.

## Skill 1: Brand Brief Keeper

Purpose: Keep Claude aligned with the brand, offer, audience, and proof boundaries.

Inputs:
- Brand notes
- priority pages
- tone rules
- competitor list
- proof points

Rules:
- Never invent credentials, clients, metrics, or case studies.
- Flag weak or unsupported claims.
- Keep one canonical brief.

Output:
- Brand brief
- do-not-say list
- claims needing proof

## Skill 2: Source Pack Builder

Purpose: Convert research into a clean evidence pack before drafting.

Rules:
- Primary sources first.
- Separate confirmed facts from interpretation.
- Write claim boundaries before the article.
- List missing sources.

Output:
- confirmed facts
- caveats
- links
- angle
- claims to avoid

## Skill 3: AI Citation Gap Finder

Purpose: Find prompts where competitors appear and the brand does not.

Rules:
- Save raw prompt answers.
- Track mentions and citations separately.
- Do not turn one run into a score.
- Label confidence.

Output:
- citation gaps
- competitor domains
- next content action

## Skill 4: Competitor Citation Teardown

Purpose: Understand what kinds of sources AI systems cite for money prompts.

Rules:
- Group by domain and format.
- Identify evidence style.
- Separate content work from link/distribution work.

Output:
- cited domains
- cited formats
- gaps
- recommendations

## Skill 5: Prompt Fan-Out Mapper

Purpose: Expand one target query into the support questions an AI answer needs.

Rules:
- Do not bloat the page.
- Group prompts by decision stage.
- Mark must-cover versus optional.

Output:
- prompt clusters
- missing sections
- FAQ candidates

## Skill 6: Content Decay Checker

Purpose: Detect and prioritize declining pages.

Rules:
- Use GSC evidence.
- Separate clicks, impressions, CTR, and position.
- Flag pages down by threshold.
- Do not invent causes.

Output:
- decay list
- lost query groups
- refresh plan

## Skill 7: Internal Link Opportunity Mapper

Purpose: Add useful internal links that help users and search engines.

Rules:
- Link because the next page helps.
- Avoid generic anchors.
- Avoid repetitive exact match anchors.

Output:
- anchor
- source placement
- target URL
- reason

## Skill 8: Schema Evidence Auditor

Purpose: Ensure structured data matches visible facts.

Rules:
- No unsupported schema.
- Validate against visible content.
- Separate error, warning, and opportunity.

Output:
- issue
- evidence
- fix
- validation

## Skill 9: Rendering And JavaScript SEO Inspector

Purpose: Compare what exists in source, render, and crawl output.

Rules:
- Do not assume client-side content is safe.
- Check links, canonicals, meta robots, schema, and primary copy.

Output:
- confirmed differences
- SEO risk
- test plan

## Skill 10: AI Answer Snapshot Logger

Purpose: Preserve raw answer evidence before summarizing.

Rules:
- Save prompt, platform, timestamp, answer, citations, and model/surface when available.
- Never rewrite the raw answer.

Output:
- answer archive
- mention/citation table

## Skill 11: Entity And Author Evidence Gate

Purpose: Confirm the page shows why the author or brand is credible.

Rules:
- Require visible experience signals.
- Link to relevant author, service, and proof pages.
- Flag generic bios.

Output:
- missing evidence
- fixes
- confidence

## Skill 12: YouTube SEO And Video Citation Mapper

Purpose: Connect article, video, description, transcript, and related videos.

Rules:
- Include article link.
- Include relevant video interlinks.
- Keep metadata natural.
- Set AI use disclosure according to the production rule.

Output:
- title
- description
- chapters
- related links

## Skill 13: Release-Day QA Gate

Purpose: Decide whether publication can proceed.

Rules:
- PASS means every blocking gate is done.
- Pending is not PASS.
- English and Spanish ship together.

Output:
- gate table
- blockers
- next action

## Skill 14: Weekly AI Visibility Report

Purpose: Turn noisy AI/search data into a weekly operating report.

Rules:
- No fake trend lines.
- Separate citations, mentions, rankings, and traffic.
- End with three priorities.

Output:
- summary
- movers
- gaps
- three priorities

## Skill 15: Memory And Decision Log Keeper

Purpose: Preserve corrections and process decisions.

Rules:
- Record user corrections.
- Record rejected assets.
- Record final approved paths.
- Record blockers with dates.

Output:
- memory note
- package ledger update
- next-action summary